The Ecomm Analyst

I open ThoughtMetric before I open Shopify. This bothers some operators when I tell them. Shopify is the source of truth for revenue. Why look at attribution first?

Because Shopify tells me what happened. Attribution tells me whether I should care.

A $14k revenue day means nothing in isolation. A $14k day from $800 in paid spend with 60% new customer rate is a great day. A $14k day from $3k in paid with 20% new customer rate is a problem dressed up as a win. Shopify shows me the first number. Attribution shows me the second and third.

The scan takes about four minutes. Here’s the order I run it.

Blended ROAS for yesterday against the trailing 7-day average. If it’s within 15%, I ignore it. If it’s outside that range, I look at why before doing anything else with the day’s numbers.

New versus returning customer split for yesterday. If new customer share dropped meaningfully, the acquisition channels are softening even when total revenue looks fine. This is the early warning that monthly numbers haven’t caught yet.

Channel mix shift. ThoughtMetric shows me first-touch attribution against survey responses. If Meta’s share of new customer acquisition jumped or dropped 10 points day over day, something changed in either the campaigns or the auction dynamics. Worth a check.

Anomalies in the survey data. When the “How did you hear about us” responses cluster oddly, like a sudden spike in “podcast” or “friend recommendation,” it usually means a piece of organic content broke through somewhere. This is how I find out about UGC moments before the operator does.

Then I open Shopify. By that point I know whether to celebrate the revenue number, ignore it, or worry about it. The number itself doesn’t change. My response to it does.

The argument for this order isn’t that attribution is more accurate than Shopify. It isn’t. The argument is that revenue is a lagging indicator and attribution is a leading one. If you only look at lagging indicators, you’re always responding to last month’s reality.